On Thu, 2004-03-25 at 06:22, Bill Stoddard wrote:
My first inclination is to say make sure the file pointer is set to 0 before you
send your brigade down the
stack. Would that solve this particular case?
However, emulate_sendfile() would still be broken in some pathological cases. For
Bojan Smojver wrote:
I think I finally found the culprit. At first I thought it was the
core_output_filter, but it turns out that emulate_sendfile (incorrectly)
assumes that it is at the beginning of the file even when it's not.
The attached patch works here when I have the combo of buckets as
Are you referring here to apr_file_seek() (which is done in
emulate_sendfile() anyway) or to tracking the file position outside?
On Tue, 2004-03-23 at 19:24, Cliff Woolley wrote:
On Mon, 22 Mar 2004, William A. Rowe, Jr. wrote:
I smell future bugs brewing - remember the handle may be dupped.
On Mon, 22 Mar 2004, William A. Rowe, Jr. wrote:
I smell future bugs brewing - remember the handle may be dupped.
Right... this is why we couldn't get away with that same trick in the
buckets code.
We tried. :)
--Cliff
On Tue, 23 Mar 2004, Bojan Smojver wrote:
Are you referring here to apr_file_seek() (which is done in
emulate_sendfile() anyway) or to tracking the file position outside?
I'm saying that file_bucket_read() does an apr_file_seek()
unconditionally, because when we tried to have it remember what
On Tue, 2004-03-23 at 20:26, Cliff Woolley wrote:
I'm saying that file_bucket_read() does an apr_file_seek()
unconditionally, because when we tried to have it remember what the
position was and bypass the seek if the position matched, file buckets
were broken whenever the apr_file_t was
I think I finally found the culprit. At first I thought it was the
core_output_filter, but it turns out that emulate_sendfile (incorrectly)
assumes that it is at the beginning of the file even when it's not.
The attached patch works here when I have the combo of buckets as
described below. The
On Tue, 23 Mar 2004, Bojan Smojver wrote:
I think I finally found the culprit. At first I thought it was the
core_output_filter, but it turns out that emulate_sendfile (incorrectly)
assumes that it is at the beginning of the file even when it's not.
The attached patch works here when I have
Bojan Smojver wrote:
I think I finally found the culprit. At first I thought it was the
core_output_filter, but it turns out that emulate_sendfile (incorrectly)
assumes that it is at the beginning of the file even when it's not.
The attached patch works here when I have the combo of buckets as
On Mon, 22 Mar 2004, Bill Stoddard wrote:
I took a 15 second look at the patch and have a concern (perhaps
unfounded). apr_file_seek() is probably an expensive operation. If
offset is 0, then in almost all cases (correct me if i'm wrong) the
fileptr will be at offset 0 as well, right? So
On Tue, 2004-03-23 at 07:48, Bill Stoddard wrote:
I took a 15 second look at the patch and have a concern (perhaps unfounded).
apr_file_seek() is probably an
expensive operation. If offset is 0, then in almost all cases (correct me if i'm
wrong) the fileptr will be at
offset 0 as well,
On Tue, 2004-03-23 at 07:48, Bill Stoddard wrote:
I took a 15 second look at the patch and have a concern (perhaps unfounded).
apr_file_seek() is probably an
expensive operation.
I've had a quick look at apr_file_seek() function and the cost depends
on how the file is opened. If it's
On Tue, 2004-03-23 at 12:40, Bojan Smojver wrote:
I did two quick benchmarks with sendfile() support turned off in order
to determine rough impact of the patch on emulate_sendfile() and the
whole of Apache (prefork MPM).
Just did another set of runs on the same machine (this time 100,000
At 03:42 PM 3/22/2004, Bojan Smojver wrote:
You are correct, it is probably an expensive operation. The other way
would be to know the position within a file and compare it to the one
that we should go to. If they are the same, do nothing.
I smell future bugs brewing - remember the handle may be
Yeah, it doesn't look good. The best I could come up with so far is the
apr_file_seek() every time. At least we know where we going to end up.
On Tue, 2004-03-23 at 14:19, William A. Rowe, Jr. wrote:
At 03:42 PM 3/22/2004, Bojan Smojver wrote:
You are correct, it is probably an expensive
15 matches
Mail list logo
